Main

Carlos A. Haro

Data scientist with experience in economic analysis

Main focus on data visualization and supervised learning.

Education

(ITAM) Instituto Tecnológico Autónomo de México

B.S., Economics

Mexico City, Mexico

2014 - 2018

Experience

Sr. Data Scientist

Mexico’s Central Tax Administration Office
(Servicio de Administración Tributaria)

Mexico City

Jan. 2019 - present

Development and deployment of a supervised model for classifying tax debt | R (tidyverse, ranger), Python (scikit-learn, pandas), SQL
Designed and taught three courses for the institution’s staff training: introduction to R, basics of exploratory data analysis, building data science pipelines using Makefiles | R (tidyverse, ggplot2), GNU Make
Designed a pipeline for automatic generation of frequent data visualization reports | R Markdown (ggplot2, shiny)
Perform network analysis to detect tax evasion communities | R (tidyverse, visnetwork, ggraph)
Version control of cloropleth maps for the geographical display of taxpayer’s data | R (leaflet QGis)

Jr. Data Scientist | Economic Analyst

EnergeA (Energy Sector Consulting Firm)

Mexico City

2018

Developed a statistical model for identifying anti-competitive practices between the mid-stream natural gas providers | R
Neighboring gas station’s competition analysis for identifying price setting mechanisms. | R
Built a PDF scrapping pipeline of 100+ files for ownership analysis of Mexico’s natural gas industry. | R (stringr, selenium)

Miscellaneous

Hackaton challenge winner

Annual BBVA Hackaton Challenge

N/A

2019

Organizer: BBVA Bank
Challenge: Update and insert operation of a 50 million observations dataset in under 10 minutes.
Result: 95%+ accuracy of update and insert achieved in 4 minutes | Pyspark on AWS for the algorithm, RMarkdown for the report

Hackaton participant

Annual Banamex Hackaton Challenge

N/A